Seagate hard drives tend to fail in their own particular ways. A faulty firmware which makes a perfectly healthy hard drive vanish from your BIOS. A shingled BarraCuda quietly rewriting itself after you've formatted it. An external Backup Plus where the bridge has failed with a standard SATA hard drive sitting behind it. On this page we go through the various ranges, the types of failure that belong to each, and what can honestly be done. Any single drive is £300 + VAT, after a free diagnostic.
Each single Seagate hard drive costs £300 + VAT, and the diagnostic before it is free. A BarraCuda out of a desktop is covered under the same price point as a FireCuda, a single IronWolf or a single external Backup Plus. Recorder disks are £400 + VAT, and a full NAS or RAID set will start at £500 + VAT.
Seagate produce all sorts of drives, ranging from the BarraCuda in a family PC through to the Exos disks in a rack, and the failure patterns are consistent enough that they are worth listing. The second is probably the one to read twice, as it is the fault most commonly mistaken for a dead disk.
The disk is spinning and sounds just as you would expect, and yet the machine behaves as though nothing is connected at all. On a Seagate this points at the service area rather than the platter surfaces, and that happens much more often than on virtually any other make. This is a true defect, one which can be repaired, and it does not mean that a single file has been lost.
A drive that reports zero bytes, a small percentage of its actual size or simply some garbage in the model string has a firmware issue and not a mechanical failure. Usually the recording surfaces will be intact. This is typically the fault that gets drives tossed out after someone has told the owner "it's dead", when it is just normal work with the right equipment.
Heads that cannot locate their position anymore, usually as a result of a knock or long-term wear. Every time it powers up from this point, the surface takes more damage. Switch it off and leave it powered down, and don't take anyone's advice about using a freezer or tapping gently on the side of the unit, as these methods will turn a recoverable drive into a scrapped one.
The usual ending for a Backup Plus or an Expansion that was unplugged in the middle of a copy. Either the partition or the file table has become damaged, and Windows is offering to do the one thing that will make sure it gets worse. Tell Windows no, unplug the device, and do not run any repair software.
Click, start, silence, and repeat, a cycle of events. On a desktop unit it is usually a matter of either the power stage or the controller board failing and not the mechanism itself, while on a portable it could be either. No diagnosis will happen by listening to the sounds coming through a wall, and this is exactly why there is a free assessment.
A copy that initially runs at full speed but quickly drops to just a few megabytes per minute is either finding sectors that can't be read cleanly, or is a shingled consumer hard disk that has filled up its staging area and is now rewriting bands in the background. Both of these need imaging hardware instead of simply a file manager and an optimistic afternoon.
This is the single biggest difference between the two major manufacturers, and it is important to know before anyone starts tearing things down. A typical Backup Plus, Expansion or One Touch has a small USB adapter board in front of a standard SATA disk, and that board does no encryption. When you remove the disk from the enclosure and plug it into a SATA port, in the normal run of things the disk should mount and show you your files. If you take the disk from a Western Digital My Passport and do the exact same thing, you will see nothing but static, as the encryption on those runs in the electronics you just removed. That contrast is set out at length on the Western Digital page.
There are three caveats on top of that. The first: Seagate does sell external models built around a password. Those units have hardware encryption like a Passport does, so a bare disk from one of them will be unreadable in just the same way. If the drive asked for a password when you plugged it in, it is one of those. Second, sometimes a shucked drive (a drive removed from its enclosure) will simply sit dead in a desktop machine because of the third pin on the SATA power connector, a reset line that enclosures never used and some power supplies hold high; external drive recovery covers that trap and the fix for it. Finally, and most importantly, none of this is a reason to open a failing drive at home. A disk that mounts is a disk being written to.
Even where the bridge does nothing clever, the board will carry the model of your drive, the revision, and occasionally a firmware quirk that will matter when the disk is put on the imager. Get a quote first, then send the entire unit, in its enclosure, with its cable and power supply. If someone has already opened up the enclosure, please put both the shell and the board in the box with the drive as well. Opening it has not ruined anything; the harm typically comes afterwards, when someone with good intentions tries to copy files off a drive that should have been imaged first.
LaCie has belonged to Seagate for over a decade, and removing the aluminium from a Rugged or a d2 will typically reveal a Seagate mechanism. That range has its own page on this site, covering the beeping, the orange bumper and the desktop units, at LaCie recovery. Seagate also bought Samsung's hard disk division years ago, so we treat a spinning Samsung SpinPoint as a Seagate; Samsung's solid state drives, however, are an entirely separate product line and nothing to do with it.
Seagate uses family names rather than colour coding, so the name of the drive tells you a good deal about the life the drive has had. A drive that has spent five years writing video streams has worn differently from one used to hold a games library, and the recovery follows that same difference.
The main desktop and laptop line, and the drive posted in to us most often. The key detail is that a good many consumer BarraCuda drives are shingled, including the widely sold 2TB models in both sizes. This means they take longer to image, and once you format one there is a clock ticking against your data instead of it waiting harmlessly until it is overwritten.
Two quite different products under one name. The older hybrids pair spinning platters with a very small flash cache on the drive's own board, which means the flash is not a separate device that can be read on its own, and if the board fails, recent writes may still be sitting in that cache. The newer FireCuda blades are NVMe solid state drives, and those are worked on the flash bench instead.
Built for NAS duty: conventional recording, vibration sensors, and designed to run continuously in a box full of its own kind. They arrive here either singly at £300 + VAT, or as a complete set out of a Synology or QNAP that has dropped a volume, in which case it is an array rebuild, priced on the number of disks, from £500 + VAT.
The surveillance line, written to constantly by DVR and NVR recorders and designed to keep streaming rather than stopping when it runs into a bad sector. They wear out over time rather than break. A disk taken out of a recorder is priced in the £400 + VAT recorder band, and recorder work is one of the stated exclusions, so it is paid upfront rather than only if we recover your data.
Rackmount capacity, in servers and storage shelves. It will often use SAS instead of SATA and will almost always be one member of a much larger set. If the whole set arrives, the job is a RAID rebuild rather than just reading one drive, and the first thing needed is the order the disks sat in and what the controller was doing at the time it stopped.
Worth naming, as many people are still asking. One generation of desktop drives shipped with firmware that could put the disk into a busy state and make it vanish from the BIOS, with every platter still intact. That particular episode is long past, but the lesson still applies today: on a Seagate, invisible does not mean finished.
Every hard disk contains a private area that you never see, where the information that allows the drive to function as itself is stored. This includes the translator, which converts the addresses a computer asks for into actual positions on the platters, the growing list of sectors it has marked as unusable, the tuning figures for each read/write head, and the model and capacity of the drive. If any of this becomes corrupted, the drive will not report a problem, since the part that would normally report it is the part that has gone. The drive simply stops answering, or answers with nonsense. To a user sitting at their desk, this looks no different from a drive that has died, and that's why so many perfectly intact Seagate drives are written off.
Reading one means talking to the disk on its own terminal and not through the operating system, rebuilding or bypassing the damaged structures, and then imaging the surfaces as normal. It is one of the more satisfying jobs on the bench because nothing is physically wrong: there is no mechanical bench work, no donor mechanism, no risk to the platters, and the whole of the data will typically come back. It is also work where in most cases you only pay if we recover your data, since it isn't an electronic or mechanical failure. A drive that has already been through several rounds of a repair utility is a harder proposition, because those tools write.
The other classic problem for Seagates is a dead controller board, usually the result of a surge through a tired mains supply or the wrong adaptor used on an external. The board can be replaced. However, replacing it is not as simple as ordering an identical board and bolting it on, which is what a depressing number of drives have had done to them before they reach us. Each board is paired to its own disk by calibration data unique to that mechanism, and that data must travel with it. On some families it sits in a small separate chip that can be moved from the old board to the new one by hand. On others it resides inside the main controller itself, which makes a straight swap useless and the job considerably more involved. Board and chip level work is one of the named exclusions, so it takes 50% upfront, with the balance only if we recover your data.
Shingled recording overlaps each track with the next so that more fits onto a platter. It is a good trade-off when most of the time the disk is being read, and an awkward one when it is being rewritten, as changing anything in the middle of a band means the whole of that band has to be rewritten. Consumer BarraCudas have used this for many years in both the 3.5-inch and the 2.5-inch ranges, and two things follow for anyone who has just lost files.
Time is the first one. Imaging one is slow and can't be hurried, so a 2TB shingled disk may take a day and a half (with everything working) where a conventional hard drive would take an afternoon. The second matters much more. After a delete or a format, the drive considers those bands free space and may write to them again while doing its own background housekeeping, which destroys the old content rather than merely unlinking it. On a conventional drive, deleted files wait patiently for someone to recover them. A shingled drive is actually eroding them while it sits there, seemingly idle and powered on. If a BarraCuda has just been formatted by mistake, switching it off within the hour is worth more than anything else you can do. Formatted drive recovery goes through what a quick format does and does not leave behind.
The drive is booked in and put on imaging hardware that can set its own timeouts and skip a difficult region instead of hammering away at it, and where needed it is assessed head by head. The report closes two working days after the drive arrives and it answers three questions: what has actually failed, how much of the disk is realistically readable, and what it will cost. The price is fixed, it is written down, and it does not change afterwards. If the report says the job is not worth doing, it will tell you so, and that costs you nothing other than the postage in.
Logical faults are worked entirely on a copy, so your drive is read but never written to. Mechanical faults go to the head-swap bench, where they are given a matched head stack from stock racked by model and firmware family, and then asked for one supervised pass before they are retired. Firmware faults are dealt with through the drive's service area without needing to open the drive at all. Regardless of the route, recovered files always come back on fresh media, and your original is either returned or securely destroyed, exactly as you instruct in writing.
Many Seagate lines come with the manufacturer's own recovery plan attached for a set number of years, examples being IronWolf Pro, the FireCuda range and some of the external models. If your drive is still within that window, and the plan was registered, then you should use it: it has already been paid for as part of the price of the drive. It is worth taking ten minutes to dig out the paperwork or the original order confirmation before ringing anybody, this laboratory included. If the plan has run out, was never registered to begin with, or your drive is not one of the covered models, the normal bands below apply.
Overwritten data is lost. A complete reformat, a reinstall onto the same volume, or weeks of continued use after the files went missing all write over the exact space being searched, and nothing can get underneath what has been written over. A shingled BarraCuda formatted a month ago and left running has been quietly rewriting itself ever since. A drive that has been opened at home has usually collected the dust that ends it. And on the FireCuda NVMe blades and other solid state products, once files are deleted from a healthy drive they typically cannot be recovered, because the controller erases freed blocks on its own within minutes; SSD recovery sets out where the line falls.
The price for a single Seagate hard drive or SSD will be £300 + VAT (including NVMe), and an external such as a Backup Plus or an Expansion is priced as a single drive. Memory cards and USB sticks are £250 + VAT. Anything that contains more than one disk, for example a NAS full of IronWolf drives or a RAID set of Exos disks, will start at £500 + VAT and goes up with the number of members. A SkyHawk or any other disk from a CCTV recorder costs £400 + VAT. In most cases you only pay if we recover your data. There are however some exceptions, namely electronic and mechanical failures, chip level work and physical damage, which is why board work and head replacement are charged at 50% of the quoted figure upfront, with the balance only if we recover your data, while DVR and forensic jobs are paid upfront. Finished single-drive jobs should be completed within 2 to 4 working days after the quote is approved.
Getting your drive to us is the easiest part of this process, once you have a quote by phone on 0800 689 0668 or through the free online diagnostic. Guildford Business Park is approximately 18 miles away, which can be covered in around a 40 minute drive via the A322 and then onto the A3, and if you would prefer to hand your device over personally rather than post it, the reception desk at the Guildford lab signs devices in Monday to Friday, 9am to 5:30pm, with no prior appointment required. In most cases, though, we receive the drives as tracked, insured parcels, posted from offices throughout the M4 corridor between Reading and Slough, from places such as Wokingham, Camberley and Ascot, and from a good deal further afield as well, as posting a box containing a disk works perfectly well from almost anywhere in the country. If the badge on yours says WD rather than Seagate, the Western Digital page covers what may differ, hard drive recovery covers every other make, and the price list publishes all five bands with VAT shown as its own line.
More Seagates arrive here with firmware faults than with mechanisms that have broken, and those give back everything. Switch it off and let the free diagnostic tell you which one you have.
Branded drives are much more likely to reach this bench in the post than to be carried in. First get a quote by telephone or with our free online diagnostic; after that, a padded parcel is better protected than something carried in a rucksack on the A322, it is covered for its value whilst it is travelling, and a drive handed across a post office counter in Bracknell on the Tuesday afternoon will generally be logged in at Guildford on the Wednesday.
An external goes in its own case. An internal comes out of the machine first. Those two rules pull in opposite directions and both of them matter. A My Passport, a My Book, a Backup Plus or an Expansion should arrive as the entire unit, screws and all, including its cable and power supply, because the electronics inside the casing may hold the only copy of the encryption key for that drive. A disk bolted into a tower, a laptop, an iMac or a recorder should arrive bare, as nothing is dismantled at this end; with a NAS, take the disks out of the chassis if you can, labelled by bay, otherwise send the NAS. Most repair shops can free one quickly. Storage soldered directly onto the mainboard is the exception with no answer: if it will not come out there is nothing to post, and if you're not sure which kind you have, ring us first.
↓ Print the shipping & booking-in form (PDF)
Address it to Guildford Data Recovery. From Bracknell the drive is approximately 18 miles (about 40 minutes) down the A322 and the A3, so once you have a quote you can drop it off in person without an appointment if you would rather. Whichever route it takes, the phone call confirming arrival comes the same day, and two working days after that the free diagnostic is finished and you have a figure.
Unsure whether yours is the sort that travels in its case? Telephone 0800 689 0668 before you tape the box, or alternatively run the free online diagnostic and answer just four questions.